Article Text

Download PDFPDF

Cardiac tissue velocities and strain rate in the early detection of myocardial dysfunction of asymptomatic boys with Duchenne’s muscular dystrophy: relationship to clinical outcome
  1. N Giatrakos1,
  2. M Kinali2,
  3. D Stephens3,
  4. D Dawson1,
  5. F Muntoni2,
  6. P Nihoyannopoulos1
  1. 1Cardiology Department, NHLI, Imperial College, London, UK
  2. 2Department of Paediatrics, Dubowitz Neuromuscular Centre, Imperial College, London, UK
  3. 3Department of Mathematics, Imperial College, London, UK
  1. Correspondence to:
    Dr Nickolaos Giatrakos
    Echocardiography Department, National Heart and Lung Institute, Imperial College, Hammersmith Hospital, Du Cane Road, London W12 ONN, UK; n.giatrakos{at}

Statistics from

Duchenne’s muscular dystrophy (DMD) is one of the most common neuromuscular disorders. Boys with DMD lose independent ambulation by the age of 12 and die of respiratory failure or cardiomyopathy in their late teens or early 20s.1

Histological changes in the heart include fibrosis, degeneration and fatty infiltration starting from the left ventricular posterior wall, which is a specific finding for DMD.2 The posterior wall can be more sensitive in identifying impaired myocardial function.

Recent consensus guidelines on the early cardiac follow up and for the treatment of asymptomatic dysfunction have been proposed.3 An unresolved issue is, however, the timing of introducing treatment. Some authors have proposed that, in view of the almost invariable development of dilated cardiomyopathy (DCM), treatment should be started even in the absence of echocardiographic signs of dysfunction. It is therefore important to be able to identify early changes that precede DCM.3

We hypothesised that strain rate (SR) can identify early myocardial dysfunction in young asymptomatic boys with DMD without conventional echocardiographic signs of DCM. We related SR to the development of cardiomyopathy over a three-year follow up.


Fifty six consecutive asymptomatic boys with DMD (mean age 8.8 (2.85) years) were enrolled. Diagnosis was confirmed by DNA analysis, unequivocal findings in muscle biopsy, or positive family history. The patients were not taking any cardiac drugs and had normal ECGs and routine echocardiograms (fractional shortening > 29%). Twenty two healthy volunteers (mean age 9 (SD 2.96) years) were recruited from the children’s outpatient department.

We used the HDI 5000 (Phillips Medical Systems) machine equipped with a P4-2 transducer for conventional and tissue Doppler imaging echocardiographic studies. The conventional study included two dimensional and Doppler blood flow velocity measurements according to standard clinical protocol. From parasternal long axis projections the colour M mode tissue Doppler imaging of the posterior wall of the left ventricle was obtained. Derived tissue velocities (TV) were analysed with dedicated research software (HDI-lab, V.1.91d, Phillips Medical Systems). The multipoint algorithm was used to trace the endocardial and epicardial borders. To reduce the random noise component of the acquired data, we defined a 0.6 mm sample volume at the subendocardium (a) and subepicardium (b).

Radial SR was defined as the difference of the instantaneous velocities (V) between the two points a and b along the ultrasound beam, divided by their distance (d) at the time of measurement4 (SR  =  Va − Vb/d).

We used the velocity gradients in time described in previous studies5 to estimate the radial SR in peak systole (SRs), in early diastole (SRed) and in late diastole (SRld). The mean TV from all points from subendocardium to subepicardium were also calculated in peak systole (TVs), early diastole (TVed) and late diastole (TVld). Corrected time to onset of relaxation (TRc) was defined as the time from the beginning of the R wave of the ECG to the transition where the SR crosses the zero line from positive to negative polarity.

Interobserver reliability was tested in 45 randomly selected cases. To test intraobserver reliability, one of the two observers re-evaluated the same 45 cases after one week. A good reproducibility of sampled data of SR, TV and TRc (< 3.5% changes for each variable) was obtained.

Patients were followed up for three years with regular clinical outpatient appointments and six-monthly conventional echocardiograms. The end point was an abnormal fractional shortening (< 29%).

Data were analysed with the SPSS package (SPSS, Chicago, Illinois, USA). Intraobserver and interobserver variabilities for SR, TV and TRc were estimated by means of absolute differences between observations. Continuous variables between groups were compared by t test for normally distributed values, as assessed by the Kolmogorov–Smirnov test. For the non-parametric statistics, we used the Mann–Whitney U test. Correlations were assessed by the Pearson’s test. The least significant difference pairwise multiple comparison test was also used. Data are presented as mean (SD).


Clinical characteristics and two dimensional and Doppler data did not differ between boys with DMD and controls. Tissue Doppler imaging values in boys with DMD were lower for TVs (26.8 (5.9) v 33.4 (11.4) mm/s, p < 0.002) and TVed (−44.7 (11.3) v −60.4 (11.9) mm/s, p < 0.000), and SRs (1.7 (0.7) v 2.8 (0.8) l/s, p < 0.000) and SRed (−5.1 (2) v −9 (2) l/s, p < 0.000) than controls, while TRc was longer (38.3 (4.1) v 36.4 (3.7) ms, p < 0.02). Multivariate analysis for heart rate, TVs, TVed, TVld, SRs, SRed, SRld and TRc differed significantly between boys with DMD and controls. On univariate analysis, there were differences in heart rate, TVs, TVed, SRs, SRed and TRc.

Thirty three of the 48 (69%) boys with DMD completed a three-year follow up (mean 40 (4) months). Seven boys (mean age 16.5 years) had deteriorated left ventricular function with fractional shortening < 29%.

SRs, SRed and SRld predicted the adverse outcome with 63.64% accuracy and TVs, TVed and TVld with 54.55% accuracy. The combination of SRed and TRc produced the highest accuracy (84.85%) of all combinations to correctly predict the poor outcome. Figure 1 (right panel) shows the predictive boundary between good and poor outcomes in terms of deterioration of cardiac function.

 Prediction of outcome according to strain rate in early diastole (SRed) and corrected time to onset of relaxation (ms) (TRc). (Top left) Receiver operating characteristic (ROC) curve with accuracy of 84.85% for SRed and TRc, which was the higher from all other combinations. (Top right) Predictive boundary between good and poor outcomes according to SRed and TRc. The patients in the red area are very likely to develop dilated cardiomyopathy. In contrast, patients in the green area will have a milder cardiac involvement. (Bottom panels) Two examples of colour M mode recording of a patient with Duchenne’s muscular dystrophy (DMD) with a poor outcome (A) and a healthy volunteer (B) (adapted to match in heart rate). We can appreciate a difference in colour in the upsloping systolic part (brighter red) and downsloping diastolic component (brighter blue).


In this study we have confirmed our original hypothesis that SR and TV may be used to identify early myocardial dysfunction in asymptomatic boys with DMD. We showed that SRs, SRed and TRc are sensitive markers for early diagnosis of myocardial dysfunction. Lastly, SR appeared to be of prognostic significance in predicting adverse outcomes in these young patients. DCM is the most common type of cardiac involvement in DMD, affecting up to 90% of boys older than 18 years. It is considered the second most common determinant of disease survival, causing 20% of deaths. The incidence rises steadily with age.1

Measurements were made at the posterior wall, which is most often affected in DMD. The mechanisms underlying the lower SR values and the prolonged TRc are uncertain but possible explanations are delayed local activation, a conduction delay or a mechanical phenomenon caused by the fibrous and fatty tissue replacement of myocardial muscular fibres.

Children with diagnosed DMD are followed up echocardiographically according to guidelines.3 Patients in our study population had normal fractional shortening and good overall systolic function but, over the period of follow up, seven developed myocardial dysfunction and three died. Current recommendations state that, if ventricular dysfunction is detected by ECG or conventional echocardiography, early treatment with angiotensin converting enzyme inhibitors, β blockers, or both should be instituted. SRed and TRc can be used to classify boys with DMD as being at high or low risk. Selectively instituting early treatment and potentially slowing down the progress to DCM may be possible. SR may be used in future longitudinal studies assessing the beneficial effect of treatment of asymptomatic boys with DMD.

Tissue Doppler imaging echocardiography may be angle dependant. Chest deformity and the tachycardia made acquisition difficult. The high signal to noise ratio may be the greatest limitation of the technique with the current technology.

In this study we showed that SR and TV are significantly lower in asymptomatic boys with DMD, when conventional echocardiography failed to show any abnormality. Myocardial function in these patients appears to be impaired from a very young age. SR and TV can be reliably used to identify early myocardial dysfunction and predict an adverse outcome. These conclusions need further validation on independent datasets and larger population samples.


Dr N Giatrakos was supported by a grant from the Hellenic Heart Foundation (ELIKAR). Dr M Kinali was partly funded by the Dutch Parents project. The ongoing support of the Muscular Dystrophy Campaign to the Dubowitz Neuromuscular Centre is gratefully acknowledged. We thank Dr Eugenio Mercuri for his valuable contribution.


View Abstract

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.